Jefferies says AI-driven data centre demand is running well ahead of physical delivery. The brokerage estimates only 8.9 GW became operational in 2025 against nearly 21.1 GW of demand, while hyperscaler capex is projected at $770bn in 2026. The signal is that capital spending alone may not resolve power, cooling, equipment and labour bottlenecks.
Publishes market analysis on AI infrastructure, data centre capacity and hyperscaler capital expenditure
Jefferies' capacity analysis gives BTW readers a quantified signal on the gap between AI infrastructure demand and deployable data centre supply.
Jefferies' capacity analysis gives BTW readers a quantified signal on the gap between AI infrastructure demand and deployable data centre supply.
The report frames AI infrastructure as a delivery-constrained market where power, cooling, equipment and labour bottlenecks can shape pricing, leasing and operator leverage.
The report frames AI infrastructure as a delivery-constrained market where power, cooling, equipment and labour bottlenecks can shape pricing, leasing and operator leverage.
Jefferies' capacity analysis gives BTW readers a quantified signal on the gap between AI infrastructure demand and deployable data centre supply.
The report frames AI infrastructure as a delivery-constrained market where power, cooling, equipment and labour bottlenecks can shape pricing, leasing and operator leverage.
| 0.90–1.00 | A | High - direct sources |
| 0.75–0.89 | A/B | Strong |
| 0.55–0.74 | B/C | Medium |
| 0.35–0.54 | C/D | Weak-medium |
| 0.10–0.34 | D | Weak signal |
| 0.00–0.09 | D | Internal monitoring |
Published reporting
• Jefferies estimates 2025 capacity supply fell short by about 12 GW See also: Microsoft.
• Cloud backlogs and equipment shortages point to persistent pricing pressure
The fact
Jefferies says AI-driven data centre demand is running well ahead of physical delivery. Its analysis estimates only 8.9 GW of capacity became operational in 2025, while demand reached nearly 21.1 GW, creating a 12 GW shortfall. Hyperscaler capex is projected at $770bn in 2026, cloud service backlogs have reached about $2tn, and the 2021-2025 cumulative undelivered capacity gap reached 20.4 GW.
The Assessment
AI infrastructure is becoming supply-constrained even as capital spending accelerates. Hyperscalers can commit more money and reserve capacity earlier, but usable supply still depends on power access, transformers, cooling systems, generators and construction labour. That strengthens operators with secured power, developers with executable sites and suppliers sitting inside the electrical and cooling bottleneck.
What to Watch
Watch whether 2026 capex converts into operational megawatts, especially in North America, where Jefferies estimates AI data centre power demand could approach 19 GW.
Signal Brief
- Signal: AI demand outruns data centre delivery
- Signal Type: AI Data Centre Supply Demand Signal
- Region: Global
- Market Class: Datacenter
Operating Surface
- Published sources should identify the affected parties, operating surface, and market exposure before this trend map is treated as complete.
Market Context
- The report frames AI infrastructure as a delivery-constrained market where power, cooling, equipment and labour bottlenecks can shape pricing, leasing and operator leverage.
- Operational relevance: High
- Time Horizon: Next quarter
What To Watch
- Watch for official statements, regulatory updates, customer or partner exposure, and follow-up disclosures.
Member Briefing
Deeper Trend Context
Sign in to unlock the full trend briefing and source notes.
Only for Strategic Circle
Strategic Circle
Open to all readers. Unlock trend briefings after joining and signing in.
Join Strategic CircleOnly for Leadership Alliance
Leadership Alliance
For operators, investors, and policy teams that need relationship evidence, failure paths, and source notes. Sign in to unlock.
Join Leadership AlliancePublic Sources and Linked Entities
13 linked-entity notes require member access.
